Exploring Substrate Binding in the Extracellular Vestibule of MhsT by

Subscriber access provided by Kaohsiung Medical University

Computational Biochemistry

Exploring substrate binding in the extracellular vestibule of MhsT by atomistic simulations and Markov models Ara M. Abramyan, Matthias Quick, Catherine Xue, Jonathan A. Javitch, and Lei Shi J. Chem. Inf. Model., Just Accepted Manuscript • DOI: 10.1021/acs.jcim.8b00175 • Publication Date (Web): 31 May 2018 Downloaded from http://pubs.acs.org on May 31, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Chemical Information and Modeling

Exploring substrate binding in the extracellular vestibule of MhsT by atomistic simulations and Markov models Ara M. Abramyan1, Matthias Quick2,4, Catherine Xue1, Jonathan A. Javitch2,3,4, Lei Shi1,* 1

Computational Chemistry and Molecular Biophysics Unit, National Institute on Drug Abuse -

Intramural Research Program, National Institutes of Health, Baltimore, Maryland, 21224, United States 2

Departments of Psychiatry and 3Pharmacology, Columbia University College of Physicians and

Surgeons, New York, New York, 10032, United States 4

Division of Molecular Therapeutics, New York State Psychiatric Institute, New York, New

York, 10032, United States

ACS Paragon Plus Environment

1

Journal of Chemical Information and Modeling 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 30

ABSTRACT

Neurotransmitter:sodium symporters (NSS) terminate neurotransmission through Na+-driven reuptake of cognate neurotransmitters. Crystallographically, whereas both substrates and inhibitors have been found to bind in the central binding (S1) site of NSS, inhibitors were found to bind to a second binding (S2) site in the extracellular vestibule (EV) of transporters for leucine (LeuT) and serotonin. Based on computational and experimental studies, we proposed that substrates bind to the S2 site of LeuT as well, and that substrate binding to the S2 site is essential for Na+-coupled symport. Recent binding experiments show that substrate (L-Trp) binding in the S2 site of MhsT, another bacterial NSS, is also central to the allosteric transport mechanism. Here, we used extensive molecular dynamics simulations combined with Markov state model analysis to investigate the interaction of L-Trp with the EV of MhsT, and identified potential binding poses of L-Trp as well as induced conformational changes in the EV. Our computational findings were validated by experimental mutagenesis studies, and shed light on the ligand binding characteristics of the EV of NSS, which may facilitate development of allosteric ligands targeting NSS.


2

Page 3 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


INTRODUCTION Neurotransmitter:sodium symporters (NSS) play an essential role in neurotransmission and include transporters for serotonin and dopamine (SERT and DAT), which are targets for antidepressants and abused psychostimulants. In a process that involves traversing between outward-facing and inward-facing conformational states, these transporters terminate neurotransmission through Na+-driven reuptake of their cognate neurotransmitters.1 The transport process can be modulated by both competitive and allosteric inhibitors. The crystal structures of NSS revealed a central binding (S1) site (Figure 1A) that binds both substrates and competitive inhibitors.2-5 For example, cocaine and benztropines bind in the S1 site of DAT, and has been found to competitively inhibit DAT.2,

6, 7

For the SERT inhibitor

citalopram, in addition to the high-affinity binding in the S1 site of SERT, a low-affinity allosteric binding site has been known to exist, and occupation of this site slows the dissociation rate of the ligand from the S1 site.8,

9

A recent crystal structure of SERT (PDB ID: 5I73)

demonstrated that the likely location of this allosteric site for S-citalopram is in the extracellular vestibule (EV) (Figure S1A).4 Interestingly, the EV also accommodates the binding pockets for several LeuT inhibitors as well (Figure S1B).5, 10-12 The substrate serotonin was shown to slow the dissociation of imipramine from SERT, suggesting the existence of an allosteric substrate binding site for serotonin in the dissociation pathway of the S1-bound imipramine.13 Computational and experimental studies in LeuT, a bacterial NSS homolog, have shown that the binding of substrate in a subpocket of the EV (termed the S2 site) triggers conformational transition towards an inward-facing state, facilitating substrate release from the S1 site.14, 15


3


Page 4 of 30

The recent crystal structures of MhsT (PDB IDs: 4US3 and 4US4),16 another bacterial NSS, were solved in an inward-occluded conformation,17 with a substrate (L-Trp)-occupied S1 site and a collapsed EV. Saturation binding studies in n-dodecyl-β-D-maltopyranoside (DDM), the detergent used for the crystallization of MhsT, supported a molar binding stoichiometry of 1. However, binding studies performed with MhsT purified in n-decyl-β-D-maltopyranoside (DM), or with MhsT reconstituted into nanodiscs showed that this NSS member features a 2:1 substrate binding stoichiometry under equilibrium conditions. Mutational analyses and site-directed thiollabeling studies reveal that the EV in MhsT indeed accommodates an S2 site that is essential for transport, just like LeuT.18 These recent findings prompted us to investigate substrate binding in the EV and the resulting conformational changes using extensive molecular dynamics (MD) simulations combined with Markov state model (MSM) analysis.


4

Page 5 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


RESULTS and DISCUSSION Based on a previously equilibrated MhsT model,17 after initial docking of a L-Trp into the EV (S2:Trp) in the presence of a L-Trp in the S1 site (S1:Trp), we carried out five rounds of MD simulations that were guided by MSM analysis after each round (see Methods). In the MSM analysis, the interacting residues of S2:Trp were identified for each frame of the MD simulations, and the identities of these residues were used as input features to build the MSM (see Methods and Figures S4 and S5). In our final MSM analysis of more than 56 µs of accumulated MD data, we identified twelve S2:Trp-bound metastable states (MSs) (Figure 1). The six MSs with the largest equilibrium probabilities account for 84% of total equilibrium probability, and are the focus of our following analysis (Figure 1B-H and Table 1).


5


Page 6 of 30

Figure 1. MSM analysis of the MD simulations identified potential S2:Trp binding sites. (A) Side view of the MhsT structure (PDB ID: 4US3) showing the relative positions of the central S1 site and the EV. (B) Equilibrium probabilities of the MSs. The six MSs with largest equilibrium probabilities are colored, whereas the other MSs are in gray. (C-H) The S2:Trp binding poses in each of the six largest MSs, with the Cα atoms of S2:Trp interacting residues


6

Page 7 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


shown in spheres (colors correspond to those in panel B), the sizes of which are proportional to their interaction frequency with S2:Trp in each MS (see Table 1). Table 1. S2:Trp-interacting residues in the six MSs with the largest equilibrium probabilities. In a MD frame, if the shortest heavy-atom distance between S2:Trp and any given residue of the protein was within 5Å, we defined that S2:Trp forms an interaction with this residue. The frequencies of interaction that are above 50% for each of the six MSs with the largest equilibrium probabilities are shaded in the corresponding MS colors. The equilibrium probability values for each MS are shown in the parenthesis.


7


Page 8 of 30

Among the six largest MSs, the S2:Trp poses in the red and orange MSs (red and orange poses) occupy more extracellular positions in the EV further away from the S1:Trp, compared to the brown, cyan, blue, and green poses (Figure S2A). Specifically, the orange pose occupies a sub-pocket between the extracellular portions of transmembrane segments (TMs) 1 and 6 (TM1e and TM6e respectively, see Figure S3 for our subsegment divisions of MhsT), whereas the red pose is in close proximity to extracellular loop 4 (EL4) (Figure 1E, H). The brown pose appears to occupy the region between EL4 and TM10e (Figure 1G), representing a transition pose between the more extracellular poses (orange and red) and the poses closer to the S1 site (cyan, blue, green). In both the cyan and blue poses, the indole ring of S2:Trp is perpendicular to the plane of membrane and its amine and carboxylic groups form salt bridge interactions with Asp385 and Arg34, respectively (Figure 1D, F). The green pose, which has the largest equilibrium probability, is distinct in that the indole ring of S2:Trp is relatively parallel to the plane of the membrane and is positioned closer to TM10e and EL5 (Figure 1C). Interestingly, we observed that the presence of S2:Trp gradually induced significant conformational changes in the EV of MhsT. In a representative trajectory shown in movie S1 in Supporting Information, the S2:Trp transitioned from the brown to the blue pose, and was then stabilized in the green pose (Figure S2B). This transition was accompanied by coordinated movements in the EV, especially near the TM10e subsegment. To quantify such changes, we calculated structural differences at both subsegment and residue levels between the most dominant green MS and the S2:apo condition (the data of which were extracted from ref

17

) using pairwise interaction analyzer for MhsT (PIA-MhsT, see Methods).

At the subsegment level, the results of our analysis indicate that the most prominent changes are the rearrangement of EL5 and TM10e on the extracellular side (Figure 2A). Indeed, when we


8

Page 9 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


further calculated the distance between TM10e and TM1e in various conditions, the large differences and fluctuations of this distance demonstrate significant flexibility of TM10e, which showed substantial conformational changes upon S2:Trp binding (Figure 3).

Figure 2. The most dominant S2-bound MS (green) is characterized by outward movements of TM10e. (A) The pairwise distances among TM subsegments in the green MS are compared to those in the S2:apo condition.17 The analysis highlights the conformational changes of TM10e and the neighboring EL5 in the green MS. (B) The pairwise Cα atom distances of the S2:Trp-interacting residues identified in the green MS (Table 1) are compared to those in the S2:apo condition. The results indicate significant changes of residues Phe366 (EL5) and Phe381 (TM10e) (see text).


9


Page 10 of 30

Figure 3. Conformational dynamics of TM10e in the S2-bound MSs. (A) A zoom-in view of the TM10e dynamics. (B) Distribution of the TM10e-TM1e distances. In both panels, the six MSs with largest equilibrium probabilities are colored in their respective colors as in Figure 1, the S2:apo condition is in black, and the crystal structure (PDB ID: 4US3) is in gray. Specifically, at the residue level, we observed significantly larger distances from Phe381 of TM10e to other S2:Trp-interacting residues in the green MS, while Phe366 of EL5 moves closer to other residues, suggesting coordinated rearrangement between these two regions upon S2:Trp binding (Figure 2B). A further analysis of the dihedral angles of these residues showed that they are highly dynamic (movie S1). In particular, in the presence of the bound S2:Trp, the sidechain of Phe381 makes a counter-clock wise rotation in the green MS from the S2:apo condition (Figure 4A, D). In the two-fold pseudosymmetry of the LeuT structural fold, TM10e is at a peripheral position from the center of the transmembrane domain and is expected to have substantial flexibility – indeed the subsegment below TM10e, TM10m, is in a bulge conformation that may allow such flexibility.19 Interestingly, TM5i and TM5m, which are pseudosymmetrical to TM10e and TM10m, undergo drastic rearrangement from an unwound to a structured configuration during conformational transitions in MhsT.16, 17 In addition, significant


10

Page 11 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


conformational changes of TM10 have been observed crystallographically in other transporters with the LeuT-like structure fold, such as Mhp1.20 How is the observed rotation of TM10e coupled to the S1 site? Compared to the S2:apo condition, we found that the changes in TM10e propagate downwards towards the S1 site through the rotamer changes of Asp385 of TM10e and Ser389 of TM10m, which are associated with the rearrangement of an S1:Trp interacting Met236 of TM8m (Figure 4E-G). Residues such as Phe381 that play a significant role in conformational transitions are expected to be critical for S2:Trp binding. Indeed, the F381C mutation in MhsT completely eliminated S2 binding and substrate transport,18 as did the mutation of the aligned residue in LeuT, Leu400.14

Figure 4. Binding of S2:Trp in the green MS is associated with conformational changes that propagate towards the S1 site. (A) Compared to the S2:apo condition (upper panel), S2:Trp


11


Page 12 of 30

binding pose in the green MS (lower panel) is associated with different orientations of residues that propagate toward the S1 site. (B-D) The changes near the S2 site are demonstrated by different distributions of sidechain rotamers of Trp33, Phe366, and Phe381 in the green MS compared to those in the S2:apo condition (in black). (E-G) The propagation of the changes from the S2 to S1 site is demonstrated by different distributions of sidechain rotamers of Asp385, Ser389, and Met236 in the green MS compared to those in the S2:apo condition (in black). The values in the crystal structure (PDB ID: 4US3) are shown in gray dotted lines. The complete elimination of S2 substrate binding by the F381C mutation is likely due both to its disruption of S2-ligand binding as well as impaired conformational transition. To further characterize the S2 substrate binding site per se, we chose Tyr37 and Phe366 for mutagenesis. These two residues interact with the S2:Trp in the highly populated MSs (Table 1) but appear not to be critical in conformational transitions (Figure 4A): specifically in the green MS, Phe366 stacks with the indole ring of S2:Trp (Figure 4A, C), while the carbonyl group of S2:Trp hydrogen bonds with Tyr37 (Figure 4A). Interestingly, the Y37C and F366C mutations resulted in biphasic binding curves (Figure 5). This observation is in stark contrast to the binding isotherm of MhsT-WT (Figure 5A), which was well fit to a rectangular hyperbola reaching a molar Trp:MhsT binding ratio of 2.18 Rectangular hyperbolic isotherms have been described for LeuT-fold proteins where both substrate binding sites exhibit similar affinities, whereas more complex (i.e., biphasic) binding isotherms have been observed when the affinities in both sites differ by a factor of about 1014, 2123

(Figure 5B, C), suggesting that in the Y37C and F366C mutants, the S2 site has a lower

affinity than that in the WT. Such disrupted S2 binding results in substantially reduced 3H-Trp


12

Page 13 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


transport as reflected in lowered initial rates and Vmax values of 3H-Trp transport as well (Figure 6).

Figure 5. Saturation equilibrium binding. 50 ng of purified protein was used with increasing concentrations (0.05–50 µM) of 3H-Trp in the presence or 150 mM NaCl. A) Fitting the data for MhsT-WT to a rectangular hyperbola model revealed an EC50 of 7.5±1.0 µM. B, C) Data fitting for Y37C and F366C was performed in two steps22: fitting of the data between 0 and 10 µM 3HTrp to the Hill equation yielded an EC50 of 3.9±0.5 µM and 6.5±2.1 µM, with Hill coefficients of 1.7±0.3 and 1.3±0.2 (indicative of allosteric interactions involving more than one binding site) for Y37C and F366C, respectively, whereas fitting the data between 10 and 50 µM 3H-Trp to a rectangular hyperbola model yielded an EC50 of 19.7±2.5 µM and 20.9±4.7 µM for Y37C and F366C, respectively. The molar binding ratio of Trp:MhsT-WT was 2.1±0.1 , while it was 2.2±0.3 and 2.3±0.4 for Y37C and F366C, respectively. D) Purified MhsT variants (1 µg protein per lane) used for the equilibrium binding experiments were subjected to 11% SDS-PAGE followed by silver staining of the gel. The approximate molecular weight of the protein standard is indicated.


13


Page 14 of 30

Figure 6. 3H-Trp uptake. (A) Time course of Trp uptake. Uptake of 0.1 µM 3H-Trp was measured in E. coli MQ614 expressing MhsT-WT (), -Y37C (), or -F366C (). MQ614 transformed with plasmid pQE60 () served as control. Data (mean ± S.E.M. of triplicate determinations) of a representative experiment (n ≥ 3) are shown. (B) Kinetics of the initial rates of 3H-Trp transport (0.1 – 50 µM). Uptake was measured for periods of 10 s (WT) or 30 s (mutants). Data (mean ± S.E.M. of triplicate determinations) of a representative experiment (n=3) were fitted with the Michaelis-Menten nonlinear regression model. (C) Relative expression of MhsT variants. A monoclonal antibody against the N-terminal His tag in the MhsT variants was used to detect relative amounts of MhsT-WT, -Y37C, or -F366C in membrane vesicles of E. coli MQ614 cells in this representative Western blot. MQ614 transformed with pQE60 served as control (co). Trp33 is a highly conserved residue among the NSS members in the EV, and was suggested to replace the role of the movable substrate in the S2 site.16 However, recent results18 indicate that, whereas Trp33 is not essential for Na+-coupled symport, a conformational arrangement that repositions its side chain may enable a Trp substrate molecule to bind in the S2 site, a critical step in the allosteric mechanism of transport. Interestingly, among the six MSs with the largest equilibrium probability, Trp33 in the green MS does not directly interact with S2:Trp, and we observed a significant change of the rotamer of Trp33 in this MS compared to the S2:apo


14

Page 15 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


condition and the crystal structure (Figure 4A, B). With all the experimental results considered, including both from our previous study18 and the Y37C and F366C results herein, we infer that the green pose is most consistent with all the experimental findings in representing the S2 binding pose. Thus, the opening of TM10e upon substrate binding appears to be an important conformational rearrangement required for the formation of the S2 binding site. As MhsT is an 11-TM transporter, its TM10e and TM10m are directly exposed to lipids. The lack of intramolecular restraints on TM10e in the transmembrane domain of 11-TM MhsT, such as those between TM10e and TM12 in the 12-TM NSS, might allow us to observe the changes in the microsecond-scale MD simulations. Nevertheless, this does not imply that this potentially functionally important rearrangement of TM10 does not happen in the 12-TM NSS.

CONCLUSIONS Taken together, by combining extensive MD simulations and MSM analysis of the substrate Trp binding to the EV of MhsT, we have identified several potential binding poses of S2:Trp. In particular, the green pose best explains all our experimental findings. However, given the complexity of the transport cycle in which MhsT has to traverse several conformational states, further investigation is required to fully reveal mechanistic consequences of S2:Trp binding. Targeting the S2 site of NSS is a promising strategy for designing allosteric modulators of transport. Our findings can be used to inform the design of new S2 compounds, as well as our understanding of the subsequent events in the transport mechanism.


15


Page 16 of 30

METHODS MD simulations. The starting MhsT model for our MD simulations was based on the S1:Trpbound crystal structure of MhsT (PDB ID: 4US4). A second Trp molecule was docked in the S2 site using the induced-fit docking (IFD) protocol24 implemented in the Schrodinger suite (release 2014-4). Visual Molecular Dynamics (VMD) program25 was used to prepare the simulation systems: the selected MhsT model from IFD was immersed in explicit POPC membrane bilayer and water solvent environment; Na+ and Cl- ions corresponding to 0.15 M concentration were added in the water phase. The energy minimization and MD simulations were carried out using NAMD26 with the CHARMM36 force field.27, 28 A cutoff of 12 Å was used for the nonbonded interactions, and long-range electrostatic effects were evaluated by the particle mesh Ewald method. The system was initially minimized for 6000 steps. In the following MD simulations, the hybrid NoseHoover Langevin piston method29 was used to maintain constant pressure at 1 atm on an anisotropic flexible periodic cell, for which a constant-ratio constraint was applied in the X-Y plane. Constant temperature at 310 K was maintained with Langevin dynamics in the isothermalisobaric (NPT) ensemble. For the first 200 ps of the first round of MD simulations (see below), a time step of 1 fs was used, which was then increased to 2 fs for the rest of the simulations. MSM analysis. We used the pyEMMA program (version 2.3.2)32 to perform the MSM analysis30, 31 as described previously,33-35 and used the binary features that describe the contacts between S2:Trp and MhsT as the input features. Specifically, if the shortest heavy-atom distance between S2:Trp and any given residue of the protein was within 5Å, the contact to that residue


16

Page 17 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


was set to 1, otherwise 0. Thus for MhsT that has 453 residues, each MD frame was represented by a 453-dimentional binary feature vector. The time-lagged independent component analysis (TICA)36, 37 was used to estimate the slow linear subspace of the input features. A dimension reduction was achieved by projecting the feature vectors on the 25 slowest TICA components (representing 80% of the cumulative kinetic variance38). k-means clustering was then used to partition the subspace into 25, 50, 100, 200, or 1000 microstates. The 50-state MSM had the highest score in terms of the variational principle,39, 40

as determined by using a previously described cross-validation protocol (Table S1).35 For this

MSM, the implied timescales (ITSs)41 for the transitions among microstates as a function of various lag times are shown in Figure S4. At each lag time, we employed the Bayesian sampling to compute statistical uncertainties of 500 transition matrix samples. Convergences of ITSs were achieved at a lag time of 120 ns, which was then consistently used for further analyses. We used the PCCA++ method42 to coarse-grain the microstates into MSs. The kinetic analysis was performed on those microstates having >60% of probability to belong to a particular MS. The final MSM was shown to be converged using Chapman-Kolmogorov test (Figure S5).31 MSM-guided iterative MD sampling. To adequately sample the S2:Trp binding modes and the conformations exhibited by MhsT in various binding modes, we used an iterative MD sampling protocol, and carried out totally 5 rounds of MD simulations based on the MSM analysis. Specifically, in the first round of MD simulations, we initiated three trajectories starting with different random seeds. By taking advantage of the MSM analysis after each round, we analyzed the population of microstates and restarted four rounds of simulations from those microstates having the smallest populations, with reducing lengths of simulations.33,


35, 43

The

17


Page 18 of 30

minimum length of trajectories was 300 ns. Overall, we collected 44 MD trajectories with an aggregated simulation time of 56.7 µs. Conformational analysis. The data sets for the conformational analysis were assembled as following: for each of the S2:Trp-bound MSs, we randomly extracted 1000 MD frames from the microstates that have >60% probability to belong to a given MS; for the equilibrated S2:apo condition, we randomly extracted 1000 frames from our previous simulation trajectories of MhsT without S2:Trp.44 We used MDTraj (version 1.7.2)45 in combination with in-house Python scripts to calculate distances and dihedral angles. We developed the pairwise interaction analyzer for MhsT (PIA-MhsT), based on our previously developed PIA-GPCR46 and PIA-DAT.35 For the analysis of coarse-grained interaction network of MhsT at subsegment level, we used the following structural elements as previously defined:17 NT (N terminus, residues 1–13), TM1i (the intracellular section (i) of TM1, 14–23), TM1m (the middle section (m) of TM1, 24–31), TM1e (the extracellular section (e) of TM1, 32–41), EL1 (42–44), TM2e (45–50), TM2m (51–60), TM2i (61–74), IL1 (the intracellular loop 1, 75–87), TM3i (88–103), TM3m (104–108), TM3e (109–124), EL2 (125– 146), TM4e (147–155), TM4i (156–164), IL2 (165–171), TM5i (172–175), TM5m (176–181), TM5e (182–194), EL3 (195–217), TM6e (218–226), TM6m (227–238), TM6i (239–245), IL3 (246–252), TM7i (253–262), TM7m (263–269), TM7e (270–282), EL4a (283–288), EL4b (289– 305), TM8e (306–319), TM8m (320–328), TM8i (329–338), IL4 (339–343), TM9i (344–353), TM9e (354–364), EL5 (365–379), TM10e (380–387), TM10m (388–392), TM10i (393–405), IL5 (406–426), and TM11 (427–448).


18

Page 19 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


In a MD frame, if the shortest heavy-atom distance between S2:Trp and any given residue of the protein was within 5Å, we defined that S2:Trp forms an interaction with this residue. The S2:Trp interacting residues for a given MS were defined to be those having frequencies of interaction with S2:Trp that are above 50% (Table 1). Site-directed mutagenesis, gene expression, protein purification. PCR-based site-directed mutagenesis was used to replace Tyr37 or Phe366 in the mhsT gene in plasmid pQO6-TEV.16 Verified mutated genes were transferred to a pNZ8048 derivative and expressed in Lactococcus lactis NZ9000 and purified to apparent homogeneity (see Fig. 5D) in 50 mM 2-amino-2(hydroxymethyl)-1,3-propanediol

(Tris)/2-(N-morpholino)-ethanesulfonic

acid

(Mes),

pH

7.5/20% glycerol/150 mM NaCl/1 mM Tris(2-carboxyethyl) phosphine (TCEP)/0.1 % (w/v) ndecyl-β-D-maltopyranoside (DM) as previously described.18,

47

Protein concentration was

measured with the amidoblack protein assay.48 Transport assay. L-[3H]tryptophan (American Radiolabeled Chemicals, Inc) transport was measured in intact E. coli MQ614 harboring pQE60 (serving as control) or pQE60 derivatives containing the mhsT-WT, -Y37C, or -F366C in buffer composed of 10 mM Tris/Mes, pH 8.5/50 mM NaCl16. Western blotting (using a monoclonal antibody against the N-terminal His tag present in all MhsT constructs) was used to determine the relative amounts of the respective MhsT variants in the membrane of MQ614 and showed that the mutations do not significantly alter the expression or insertion of the protein in the E. coli plasma membrane (Fig. 6C). Membrane vesicles were produced from MQ614 cells expressing the given mhsT variants or those transformed with pQE60 (no mhsT). 10 µg of total membrane protein were subjected to SDS-PAGE followed by immunoblot analysis using His probe (Santa Cruz Biotechnology, Inc.)


19


Page 20 of 30

and horseradish peroxidase-based chemiluminescence detection (SuperSignal® West Pico kit, Thermo Scientific). Scintillation proximity assay (SPA)-based binding studies. SPA-based L-[3H]tryptophan binding to purified MhsT was performed in 50 mM Tris/Mes, pH 7.5/20% glycerol/150 mM NaCl/1 mM TCEP/0.1% DM or DDM. 50 ng protein were bound to Cu2+-coated yttrium silicate (YSi) SPA beads (2.5 mg/mL) in 100 µL in the dark for 16 h at 4 ºC with vigorous shaking. The binding activity was determined with a Wallac 1450 MicroBetaTM plate PMT counter in the SPA mode. The non-proximity background signal was determined in the presence of 800 mM imidazole to prevent the interaction of the His-tagged protein with the Cu2+-coated SPA beads. Thus, specific binding was determined by subtracting the nonspecific binding from the total binding. Nonlinear regression fitting of the data (mean ± S.E.M. of triplicate determinations) of a representative experiment (n≥2) was performed in Prism 7 to obtain the EC50 and molar ratios of ligand-toMhsT binding by plotting specific binding as function of free Trp. The binding isotherm for MhsT-WT was determined by fitting the data to a one-site binding model, whereas data fits of the two phases observed for MhsT-Y37C and -F366C were performed independently with onesite models. Data analysis. All experiments were performed at least in duplicate with replicas of ≥ 3 and data are expressed as mean ± S.E.M.


20

Page 21 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ASSOCIATED CONTENT Supporting Information. The Supporting Information is available free of charge on the ACS Publications website. Table S1: Variational scores for a range of microstates. Figure S1: The S1 and S2 sites in hSERT and LeuT. Figure S2: Distances between S1:Trp and S2:Trp. Figures S3: MhsT subsegment division. Figure S4: Implied timescales for the 50-microstate MSM. Figure S5: ChapmanKolmogorov test of the MSM. (PDF) Movie S1: The stabilization of S2:Trp in the green pose is associated with conformational changes near TM10e of MhsT. (MPG)

AUTHOR INFORMATION Corresponding Author * To whom correspondence should be addressed: Lei Shi, 333 Cassell Drive, Room 1121, Baltimore, MD 21224, (443)740-2774, [email protected]. Author Contributions A.M.A., M.Q., J.A.J., and L.S. designed the study. A.M.A., C.X., and L.S. carried out the computations and analysis. M.Q. carried out experimental work. All the authors took part in interpreting the results. A.M.A., M.Q., and L.S. wrote the initial draft, with all authors participating in revising and finalizing the manuscript. Funding Sources


21


Page 22 of 30

Support for this research was provided by the Intramural Research Program of National Institutes of Health (NIH), National Institute on Drug Abuse (A.M.A., C.X., and L.S.) and by NIH grants DA041510 and GM087519 (M.Q. and J.A.J.).

ACKNOWLEDGMENT This work utilized the computational resources of the NIH HPC Biowulf cluster (http://hpc.nih.gov).


22

Page 23 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


REFERENCES 1.

Kristensen, A. S.; Andersen, J.; Jorgensen, T. N.; Sorensen, L.; Eriksen, J.; Loland, C. J.;

Stromgaard, K.; Gether, U., SLC6 Neurotransmitter Transporters: Structure, Function, and Regulation. Pharmacol. Rev. 2011, 63, 585-640. 2.

Wang, K. H.; Penmatsa, A.; Gouaux, E., Neurotransmitter and Psychostimulant

Recognition by the Dopamine Transporter. Nature 2015, 521, 322-327. 3.

Yamashita, A.; Singh, S. K.; Kawate, T.; Jin, Y.; Gouaux, E., Crystal Structure of a

Bacterial Homologue of Na+/Cl--dependent Neurotransmitter Transporters. Nature 2005, 437, 215-223. 4.

Coleman, J. A.; Green, E. M.; Gouaux, E., X-ray Structures and Mechanism of the

Human Serotonin Transporter. Nature 2016, 532, 334-339. 5.

Singh, S. K.; Piscitelli, C. L.; Yamashita, A.; Gouaux, E., A Competitive Inhibitor Traps

LeuT in an Open-to-out Conformation. Science 2008, 322, 1655-1661. 6.

Beuming, T.; Kniazeff, J.; Bergmann, M. L.; Shi, L.; Gracia, L.; Raniszewska, K.;

Newman, A. H.; Javitch, J. A.; Weinstein, H.; Gether, U.; Loland, C. J., The Binding Sites for Cocaine and Dopamine in the Dopamine Transporter Overlap. Nat. Neurosci. 2008, 11, 780-789. 7.

Bisgaard, H.; Larsen, M. A.; Mazier, S.; Beuming, T.; Newman, A. H.; Weinstein, H.;

Shi, L.; Loland, C. J.; Gether, U., The Binding Sites for Benztropines and Dopamine in the Dopamine Transporter Overlap. Neuropharmacology 2011, 60, 182-190.


23


8.

Page 24 of 30

Plenge, P.; Mellerup, E. T., Antidepressive Drugs can Change the Affinity of

[3H]Imipramine and [3H]Paroxetine Binding to Platelet and Neuronal Membranes. Eur. J. Pharmacol. 1985, 119, 1-8. 9.

Plenge, P.; Shi, L.; Beuming, T.; Te, J.; Newman, A. H.; Weinstein, H.; Gether, U.;

Loland, C. J., Steric Hindrance Mutagenesis in the Conserved Extracellular Vestibule Impedes Allosteric Binding of Antidepressants to the Serotonin Transporter. J. Biol. Chem. 2012, 287, 39316-39326. 10. Singh, S. K.; Yamashita, A.; Gouaux, E., Antidepressant Binding Site in a Bacterial Homologue of Neurotransmitter Transporters. Nature 2007, 448, 952-956. 11. Zhou, Z.; Zhen, J.; Karpowich, N. K.; Law, C. J.; Reith, M. E.; Wang, D. N., Antidepressant Specificity of Serotonin Transporter Suggested by Three LeuT-SSRI Structures. Nat. Struct. Mol. Biol. 2009, 16, 652-657. 12. Zhou, Z.; Zhen, J.; Karpowich, N. K.; Goetz, R. M.; Law, C. J.; Reith, M. E.; Wang, D. N., LeuT-desipramine Structure Reveals How Antidepressants Block Neurotransmitter Reuptake. Science 2007, 317, 1390-1393. 13. Wennogle, L. P.; Meyerson, L. R., Serotonin Modulates the Dissociation of [3H]Imipramine from Human Platelet Recognition Sites. Eur. J. Pharmacol. 1982, 86, 303-7. 14. Shi, L.; Quick, M.; Zhao, Y.; Weinstein, H.; Javitch, J. A., The Mechanism of a Neurotransmitter:Sodium Symporter--Inward Release of Na+ and Substrate is Triggered by Substrate in a Second Binding Site. Mol. Cell 2008, 30, 667-677.


24

Page 25 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


15. Zhao, Y.; Terry, D. S.; Shi, L.; Quick, M.; Weinstein, H.; Blanchard, S. C.; Javitch, J. A., Substrate-Modulated Gating Dynamics in a Na+-coupled Neurotransmitter Transporter Homologue. Nature 2011, 474, 109-113. 16. Malinauskaite, L.; Quick, M.; Reinhard, L.; Lyons, J. A.; Yano, H.; Javitch, J. A.; Nissen, P., A Mechanism for Intracellular Release of Na+ by Neurotransmitter/Sodium Symporters. Nat. Struct. Mol. Biol. 2014, 21, 1006-1012. 17. Stolzenberg, S.; Li, Z.; Quick, M.; Malinauskaite, L.; Nissen, P.; Weinstein, H.; Javitch, J. A.; Shi, L., The role of Transmembrane Segment 5 (TM5) in Na2 Release and the Conformational Transition of Neurotransmitter:Sodium Symporters toward the Inward-open State. J. Biol. Chem. 2017, 292, 7372-7384. 18. Quick, M.; Abramyan, A. M.; Wiriyasermkul, P.; Weinstein, H.; Shi, L.; Javitch, J. A., The LeuT-fold Neurotransmitter:Sodium Symporter MhsT has Two Substrate Sites. unpublished results. 19. Cartailler, J. P.; Luecke, H., Structural and Functional Characterization of π Bulges and Other Short Intrahelical Deformations. Structure 2004, 12, 133-144. 20. Weyand, S.; Shimamura, T.; Yajima, S.; Suzuki, S.; Mirza, O.; Krusong, K.; Carpenter, E. P.; Rutherford, N. G.; Hadden, J. M.; O'Reilly, J.; Ma, P.; Saidijam, M.; Patching, S. G.; Hope, R. J.; Norbertczak, H. T.; Roach, P. C.; Iwata, S.; Henderson, P. J.; Cameron, A. D., Structure and molecular mechanism of a nucleobase-cation-symport-1 family transporter. Science 2008, 322, 709-713.


25


Page 26 of 30

21. Quick, M.; Winther, A. M.; Shi, L.; Nissen, P.; Weinstein, H.; Javitch, J. A., Binding of an Octylglucoside Detergent Molecule in the Second Substrate (S2) Site of LeuT Establishes an Inhibitor-bound Conformation. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 5563-5568. 22. Li, Z.; Lee, A. S.; Bracher, S.; Jung, H.; Paz, A.; Kumar, J. P.; Abramson, J.; Quick, M.; Shi, L., Identification of a Second Substrate-Binding Site in Solute-Sodium Symporters. J. Biol. Chem. 2015, 290, 127-141. 23. Zehnpfennig, B.; Wiriyasermkul, P.; Carlson, D. A.; Quick, M., Interaction of α-Lipoic Acid with the Human Na+/Multivitamin Transporter (hSMVT). J. Biol. Chem. 2015, 290, 1637216382. 24. Sherman, W.; Day, T.; Jacobson, M. P.; Friesner, R. A.; Farid, R., Novel Procedure for Modeling Ligand/Receptor Induced Fit Effects. J. Med. Chem. 2006, 49, 534-553. 25. Humphrey, W.; Dalke, A.; Schulten, K., VMD: Visual Molecular Dynamics. J. Mol. Graphics 1996, 14, 33-38. 26. Phillips, J. C.; Braun, R.; Wang, W.; Gumbart, J.; Tajkhorshid, E.; Villa, E.; Chipot, C.; Skeel, R. D.; Kale, L.; Schulten, K., Scalable Molecular Dynamics with NAMD. J. Comput. Chem. 2005, 26, 1781-1802. 27. MacKerell, A. D.; Bashford, D.; Bellott; Dunbrack, R. L.; Evanseck, J. D.; Field, M. J.; Fischer, S.; Gao, J.; Guo, H.; Ha, S.; Joseph-McCarthy, D.; Kuchnir, L.; Kuczera, K.; Lau, F. T. K.; Mattos, C.; Michnick, S.; Ngo, T.; Nguyen, D. T.; Prodhom, B.; Reiher, W. E.; Roux, B.; Schlenkrich, M.; Smith, J. C.; Stote, R.; Straub, J.; Watanabe, M.; Wiórkiewicz-Kuczera, J.; Yin,


26

Page 27 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


D.; Karplus, M., All-Atom Empirical Potential for Molecular Modeling and Dynamics Studies of Proteins†. J. Phys. Chem. B 1998, 102, 3586-3616. 28. Best, R. B.; Zhu, X.; Shim, J.; Lopes, P. E.; Mittal, J.; Feig, M.; Mackerell, A. D., Jr., Optimization of the Additive CHARMM All-atom Protein Force Field Targeting Improved Sampling of the Backbone Φ, Ψ and Side-Chain Χ1 and Χ2 Dihedral Angles. J. Chem. Theory Comput. 2012, 8, 3257-3273. 29. Feller, S. E.; Zhang, Y.; Pastor, R. W.; Brooks, B. R., Constant Pressure Molecular Dynamics Simulation: The Langevin Piston Method. J. Chem. Phys. 1995, 103, 4613-4621. 30. Noé, F.; Horenko, I.; Schütte, C.; Smith, J. C., Hierarchical Analysis of Conformational Dynamics in Biomolecules: Transition Networks of Metastable States. J. Chem. Phys. 2007, 126, 04B617. 31. Prinz, J. H.; Wu, H.; Sarich, M.; Keller, B.; Senne, M.; Held, M.; Chodera, J. D.; Schutte, C.; Noe, F., Markov Models of Molecular Kinetics: Generation and Validation. J. Chem. Phys. 2011, 134, 174105. 32. Scherer, M. K.; Trendelkamp-Schroer, B.; Paul, F.; Perez-Hernandez, G.; Hoffmann, M.; Plattner, N.; Wehmeyer, C.; Prinz, J. H.; Noe, F., PyEMMA 2: A Software Package for Estimation, Validation, and Analysis of Markov Models. J. Chem. Theory Comput. 2015, 11, 5525-5542. 33. Plattner, N.; Noe, F., Protein Conformational Plasticity and Complex Ligand-Binding Kinetics Explored by Atomistic Simulations and Markov Models. Nat. Commun. 2015, 6, 7653.


27


Page 28 of 30

34. Wieczorek, M.; Sticht, J.; Stolzenberg, S.; Gunther, S.; Wehmeyer, C.; El Habre, Z.; Alvaro-Benito, M.; Noe, F.; Freund, C., MHC Class II Complexes Sample Intermediate States along the Peptide Exchange Pathway. Nat. Commun. 2016, 7, 13224. 35. Abramyan, A. M.; Stolzenberg, S.; Li, Z.; Loland, C. J.; Noe, F.; Shi, L., The Isomeric Preference of an Atypical Dopamine Transporter Inhibitor Contributes to Its Selection of the Transporter Conformation. ACS Chem. Neurosci. 2017, 8, 1735-1746. 36. Perez-Hernandez, G.; Paul, F.; Giorgino, T.; De Fabritiis, G.; Noe, F., Identification of Slow Molecular Order Parameters for Markov Model Construction. J. Chem. Phys. 2013, 139, 015102. 37. Schwantes, C. R.; Pande, V. S., Improvements in Markov State Model Construction Reveal Many Non-Native Interactions in the Folding of NTL9. J. Chem. Theory Comput. 2013, 9, 2000-2009. 38. Noe, F.; Clementi, C., Kinetic Distance and Kinetic Maps from Molecular Dynamics Simulation. J. Chem. Theory Comput. 2015, 11, 5002-5011. 39. Noe, F.; Nuske, F., A Variational Approach to Modeling Slow Processes in Stochastic Dynamical Systems. Multiscale Model. Simul. 2013, 11, 635-655. 40. Nuske, F.; Keller, B. G.; Perez-Hernandez, G.; Mey, A. S.; Noe, F., Variational Approach to Molecular Kinetics. J. Chem. Theory Comput. 2014, 10, 1739-1752. 41. Swope, W. C.; Pitera, J. W.; Suits, F., Describing Protein Folding Kinetics by Molecular Dynamics Simulations. 1. Theory. J. Phys. Chem. B 2004, 108, 6571-6581.


28

Page 29 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


42. Roblitz, S.; Weber, M., Fuzzy Spectral Clustering by PCCA Plus : Application to Markov State Models and Data Classification. Advances in Data Analysis and Classification 2013, 7, 147-179. 43. Verma, R. K.; Abramyan, A. M.; Michino, M.; Free, R. B.; Sibley, D. R.; Javitch, J. A.; Lane, J. R.; Shi, L., The E2.65A Mutation Disrupts Dynamic Binding Poses of SB269652 at the Dopamine D2 and D3 Receptors. PLoS Comput. Biol. 2018, 14, e1005948. 44. Stolzenberg, S.; Quick, M.; Zhao, C.; Gotfryd, K.; Khelashvili, G.; Gether, U.; Loland, C. J.; Javitch, J. A.; Noskov, S.; Weinstein, H.; Shi, L., Mechanism of the Association between Na+ Binding and Conformations at the Intracellular Gate in Neurotransmitter:Sodium Symporters. J. Biol. Chem. 2015, 290, 13992-14003. 45. McGibbon, R. T.; Beauchamp, K. A.; Harrigan, M. P.; Klein, C.; Swails, J. M.; Hernandez, C. X.; Schwantes, C. R.; Wang, L. P.; Lane, T. J.; Pande, V. S., MDTraj: A Modern Open Library for the Analysis of Molecular Dynamics Trajectories. Biophys. J. 2015, 109, 15281532. 46. Michino, M.; Boateng, C. A.; Donthamsetti, P.; Yano, H.; Bakare, O. M.; Bonifazi, A.; Ellenberger, M. P.; Keck, T. M.; Kumar, V.; Zhu, C.; Verma, R.; Deschamps, J. R.; Javitch, J. A.; Newman, A. H.; Shi, L., Toward Understanding the Structural Basis of Partial Agonism at the Dopamine D3 Receptor. J. Med. Chem. 2017, 60, 580-593. 47. Quick, M.; Javitch, J. A., Monitoring the Function of Membrane Transport Proteins in Detergent-Solubilized Form. Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 3603-3608.


29


Page 30 of 30

48. Schaffner, W.; Weissmann, C., A Rapid, Sensitive, and Specific Method for the Determination of Protein in Dilute Solution. Anal. Biochem. 1973, 56, 502-514.

For Table of Contents Use Only

Exploring substrate binding in the extracellular vestibule of MhsT by atomistic simulations and Markov models Ara M. Abramyan, Matthias Quick, Catherine Xue, Jonathan A. Javitch, Lei Shi*


30

Exploring Substrate Binding in the Extracellular Vestibule of MhsT by

Recommend Documents